Federated Learning Cohorts (FLoC) is Google's mechanism to protect individuals from identification by unscrupulous internet marketers, without reaping the full benefits of targeted advertising. However, the idea of grouping users into a cohort has not been well received by privacy advocates. In fact, the Electronic Frontier Foundation (EFF) believes that FLoC is Google's attempt to create a replacement for the third-party cookie. As Google begins to roll out and test FLoC in limited markets around the world, TechRadar Pro spoke with Gowthaman "G'man" Ragothaman, CEO of Aqilliz, to understand if FLoC is really as bad as it is," Aqilliz said. , based in Singapore, uses a blockchain-based solution to solve some of the same problems as Google FLoC without breaking privacy laws, such as the European General Data Protection Regulation (GDPR).
Could you help us get over Google's marketing hyperbole and summarize what exactly FLoC is and why it was designed?
To understand Google's Federated Learning Cohort (FLoC), we need to go back a bit. In early 2020, Google announced that it would phase out support for third-party cookies in its Chrome browser by 2022, following in the footsteps of Apple's Safari and Mozilla's Firefox browsers. The impact of this, of course, is significant because it marks the decline in support for one of the most ubiquitous user tracking techniques on the Internet. Without third-party cookies in these three browsers, approximately 87% of online users can no longer be tracked using this method. Arguably, Google's strategy cannot be blamed at first glance. The move is a step in the right direction in the eyes of privacy advocates (for the most part), and sets a new precedent for the future of user credentials. To make sure advertisers don't find themselves without tools at their disposal, Google has been working on its Privacy Sandbox since 2019 to push for a more privacy-focused Internet. This sandbox aims to set open standards for tracking users while protecting their right to privacy and FLoC is part of one of the sandbox initiatives. FLoC allows the creation of targeted advertising and audience profiles without collecting a person's browsing history. Instead, FLoC allows advertisers and AdTech companies to target a "cohort" (or "herd") of similar users. Browsers develop, update, and refine these cohorts based on users' browsing behavior. Those with similar browsing habits are grouped together and these cohorts are updated over time as users browse the web. By leveraging machine learning algorithms, the browser can build a herd based on specific input features (for example, the URL of visited sites or page content). FLoC is unique in that it ensures that all browser history remains in local storage; the browser only "reveals" the generated herd that the user is a part of. However, the success of FLoC depends on the quality of these groups. Labels must be suitable for machine learning, and clusters must be meaningfully distributed and sized so that they accurately represent the interests of the group as a whole, rather than an individual. While real-world testing of FLoC has already begun, Google recently announced that it might not work in countries where GDPR and the Electronic Privacy Directive are in effect due to privacy concerns. While this is an alternative that is supposed to offer more privacy to users, some regulators argue that it is still not enough.
Even as Google begins testing FLoC, it faces opposition from various quarters. There are antitrust lawsuits and privacy advocates believe this is just a disguised replacement for third-party cookies. Is FLoC really as bad as claimed?
Let's not forget that FLoC is just one of many solutions offered in today's advertising ecosystem to address the end of third-party cookies. From The Trade Desk Unified Identifier 2.0 to our own solution on Aqilliz, the industry is mitigating the current existential threat it now faces as it envisions a future without third-party credentials. I still wouldn't fully condemn FLoC; With testing ongoing, I'm sure we'll definitely see more improvements to Google's algorithm. This year, and beyond, should be a time of trial and error as well as experimentation.
Does the reaction against FLoC mean that we are intended to be tracked by third-party cookies? Do you think cohort targeting is a viable solution if approached in a different way than Google?
The world has changed a lot since the appearance of third-party cookies on the web. The cookies were free, open source, and had no liability on behalf of digital supply chain participants regarding consumer preferences or consent. With consumer consent now at the top of the privacy priority list, expectations have changed. From large-scale data breaches to incredibly accurate ads, consumers and regulators expect to see a universal identifier that can capture and transmit consumer consent throughout the supply chain. In light of this, I think Google's decision to move to aggregating information via cohorts instead of user-level identifiers is a step in the right direction. The reality is that independent identification at the user level will have to give way to communication at the cohort level for much advertising, unless and until an individual's explicit consent is given to receive promotional communications and allow the online tracking.
How are blockchains used for online advertising? Based on your experience using the technology at Aqilliz, what are the benefits of using the type of blockchain you are using, both from a user perspective and from the perspective of an online advertiser?
At Aqilliz, blockchain is used to guarantee three main functionalities in all our solutions: transparency, traceability and immutability. All of these correspond to existing problems in the AdTech and MarTech ecosystems. From a user perspective, we can ensure that all data is handled appropriately (anonymized, grouped, and aggregated) to ensure that it cannot be reverse-engineered to reveal individual data points. We may maintain and protect privacy in accordance with local data protection laws where our customers are located. This means that users can still benefit from properly tailored marketing messages, and brands can always maximize their spend to ensure they reach the right audiences at the right time. In addition to data privacy compliance, another area we look at is real-time campaign optimization. With the number of intermediaries involved in today's opaque digital campaign supply chain, it's difficult to optimize campaign metrics in real time. Savings of time and money are lost due to the amount of data that must be ingested and reconciled later. Our blockchain-based in-flight campaign optimization solution was designed to add transparency and accountability to the programmatic supply chain while providing real-time visibility. Smart contracts enable built-in content authentication through pre-coded requirements, such as viewability thresholds, and automate agreements to ensure only verifiable, fraud-free, and brand-safe ads are paid, saving potential dollar savings advertising, a great advantage when budgets are tight.
What about strict data compliance laws like GDPR? FLoC does not seem to pass the limit. Why do you think a distributed ledger based solution will do this?
Distributed ledger technology can manifest itself in many different forms and at Aqilliz we have identified that a hybrid solution that combines the best of what public and private blockchains have to offer works best. Our underlying technology infrastructure, Atom, implements the best privacy proxy techniques, such as cryptography and differential privacy, to enable brands to discover new audiences on the open web in a federated manner. During this time, we have combined the security and scalability of a private blockchain with the accountability and decentralization of a public blockchain. In accordance with data compliance laws like GDPR, our solutions do not reveal individual raw data; instead, they remain in local storage (either on user devices or on brand-owned servers) before being anonymized and then treated with differential confidentiality. Differential privacy masks individual data points with statistical noise so that only aggregated information can be revealed without any possibility of reverse engineering the data set. In the meantime, we can also provide an independent and immutable record of the processing of any inter-company activity, in accordance with the GDPR clause for a record of processing activities. As you can see, distributed ledger technology is not a one-size-fits-all solution that can be used out of the box. Using our work at Aqilliz as an example, distributed ledger technology needs to be complemented by other enhancements to provide a holistic, privacy-compliant infrastructure.
He spoke of the inefficiencies of the marketing technology ecosystem as it is today. Could you share some of these concerns and how do you think they can be overcome without compromising user privacy?
The inefficiencies that occur in the MarTech ecosystem are more detrimental to brands and advertisers than to users. Whether due to a lack of transparency, a lack of standardization across measurement standards, or the presence of fraud, these pain points have a ripple effect on users because they have a major impact on the extent to which brands can successfully tailor their marketing messages. marketing. in the reality. ...